Developing a Common Model for Predicting the Wax Deposition in Crude Pipelines

Wax Deposition commonly occurs in hot waxy crude oil pipelines. How to precisely predict the wax deposition is significant to the safe operation of pipelines. Presently the applied way to predict wax deposition in crude pipelines is mainly through numerical solutions by using the wax deposition mathematical model aimed at pipelined crude and relevant pipeline parameters, in which creating a precise wax deposition model plays an important role. Because wax deposition strongly depends on the composition of crude, most parameters in wax deposition model have to be determined through laboratory loop tests. Previous practices show that loop tests not only take a long time but also consume a large amount of crude sample. To obtain a precise model all the above had to be done in the past. Many wax deposition loop tests aimed at different crude oils have been carried out in our laboratory and, more importantly, a great deal of test data have been collected and analyzed. Now, based on our new research results on mechanism of wax deposition in crude pipelines and test data of up to 9 representative crudes, a common rule about wax deposition available to most crudes and their tests on loop has been excitedly discovered. It has resulted in the development of a new practical and more efficient mathematical model, which no longer needs long-time loop tests and large amount of crude sample and could be commonly or generally used on wax deposition prediction of most crude oils in pipelines. In this paper, how to discover the common rule about wax deposition of most crudes is introduced in detail. Also, a verification case for the application of the new model on a practical crude pipeline in China has been implemented by comparing the prediction results according to the new model with the operational data from the field during 7 pigging cycles. It is necessary to emphasize that the crude selected in this case for verifying is not among the 9 crudes used to establish the new model. The verification results show that the average error of prediction using the new model is only −4.215% and the maximal error is −20%, which are far better than ±30% by value in average prediction error tolerated on engineering application of wax deposition prediction around the world currently. So the results are very successful and encouraging and imply a prospective application.